My Marconi 702 has a strange behavior at turnon.
When it warms up, the screen is very slow to fill up, spreading out from the center. This is not strange, its
reported by lots of owners of these sets.
What's strange is the behavior of the brightness control. Just after the screen fills up, its still a bit dim
unless the brightness is all the way up, when its fairly much normal. But as I turn the brightness down,
its very very strange.
As it gets turned down, the picture washes out and essentially disappears into a gray screen. As I turn
it down more I would have expected the picture to either go negative or just get darker and darker and
then go black. But what it does do is reappear as a regular positive image, with the brightest parts
getting darker very very slowly and the darkest parts soon going quite dark.
As the set warms up for another 10 minutes this behavior goes away, and at brightness settings lower
than the previous "contrast goes to zero" settings the picture does just get darker quickly.
What could cause this? I was thinking a gassy CRT, but I see no sign of a glow in it. I was thinking a
bad contact on a CRT pin or perhaps the detector diode. When I first got the set there was a bad contact,
but a good cleaning seems to have made it stable. The set has plenty of drive to the CRT. The actual
picture has a perfectly normal appearance with good brightness and contrast.
Another symptom is that one corner of the picture is somewhat out of focus. I've tried magnets to
fix that, thinking maybe a magnetized housing that a strong degausser can't fix, but they don't help.
I'd like opinions first before I go into it with my scope. I would have only a few minutes a day to try as it needs
to cool all the way down before another restart.
I hope this is the right section for this query.
Your description does not mention of the sensitivity control, which would have a big impact on the screen using a direct feed from an Aurora etc.
That said, if you have adjusted this in addition to the contrast control then next would start by measuring both the cathode and grid voltages of the CRT using analogue meters observing the raster. This should show if the CRT or the drive is the issue.
Please disregard if this is all obvious to you...
taken from 702 user manual
The sensitivity control on my set does little. Somebody changed the circuit to reduce its range. I adjust
"sensitivity" with an attenuator on the input. I have it set so with sensitivity all the way up its just below
the compression level when the output from the RF unit is observed on a scope.
What you suggest is almost what I propose to do next, except I'll use a dual channel DC scope.
I hooked up a scope and turned it on. It behaved perfectly normally, no sign of the odd behavior.
I did notice that a few parts, mostly in the sync unit, might have been being pushed on
by the side panel. I tried gently moving them away from it enough so they would not touch it.
No effect at all. Still normal. I removed the scope probes. Still normal. I put the sides on.
Still normal.
This is the sort of thing I have learned to expect.
Watch the brightness of the CRT heater as things happen. Poor CRT pin connections can be a problem especially for the heater. You could attach wire stubs to the heater pins to let you monitor them.
Peter
Peter: I thought of that the first time it happened and did look through the screened back.
My set has a replacement back that's fairly open, and one can sort of tell about the CRT.
By the time I got as good a look as I could, it looked quite OK. I tried again yesterday,
but the problem did not happen.
Does anybody know what the official frequencies for the Test Card C bars with 4, 6, 8, 10,
and 12 bars are supposed to be? I'm trying to get test patterns that are not "converted"
from NTSC but generated directly by my HackRF gizmo and can of course look on a scope
to check. The idea is working nicely and I can get up to about 5 Mhz generated lines
with the clock at 12.5 MHz with no aliasing. But I don't know what the pixel clock is
supposed to be. The problem is that the HackRF has a limited set of clock frequencies for
baseband. I'm currently running at 52.8 Hz vertical.
P.S.: I was told my set came from Bennett-Levy. But unless somebody changed the veneer
its not the one in the big auction which I was looking at yesterday. You've got it on your list.
P.P.S. The HackRF works well for transmitting but "hack" is a good description ... your have
to write software to generate your patterns and what it does is not really well documented.
OTOH, if you need 5 gHz out its the bee's knees. But I'm a born hacker. Its FUN. The
HackRF is bad bad bad for receiving, poor dynamic range.
The frequencies for the bars inside the test card C are:
4 bars: 1 MHz
6 bars: 1.5 MHz
8 bars: 2 MHz
10 bars: 2.5 MHz
12 bars: 3 MHz.
Hope this helps.
Best regards, Jac
Thanks for the frequencies. It turns out that my calculations were
correct. It turns out the frequency response to the CRT grid is
flat to 2 MHz, is down about 4 or 5dB at 2.5 and there is still good
response at 3 MHZ. I've tuned the set to be VSB lower sideband.
This was pretty much determined by the fact that the core of the
3rd if transformer is frozen in place.
Its time for the Super Bowl, which I'm going to watch on my 1954
color TV. So I turned the Marconi off and tried to close the lid.
That failed!
It was stuck. Removing the screw connecting the apparatus to the lid,
the apparatus was still stuck. Really stuck. Trying to move it wiggled
the RF chassis a bit. With some grease I got it loose but after further
greasing it didn't hold well. It had been getting worse and worse over
time to the point that the lid was in dire danger of falling. A bit of
filing at one point got it to hold precariously but when lowering it stuck
again. It turns out both the "knife" and the post it rubs on are too
worn to fix so new ones would need to be fabricated. I could do this
easily as I have access to a full machine shop. But ... I see no way
to get the mechanism out without removing the whole RF unit.
So ... since the knife is currently well and safely stuck so that the
lid will close I'm going to make a small metal bar that will pivot off
the stuck mechanism's top hole and attach to the lid's metal
bracket with a cotter pin when in use, and I'll pull that out to
stow the lid.
At least the Super Bowl seems to be an exciting game.
P.S.: I was told my set came from Bennett-Levy. But unless somebody changed the veneer
its not the one in the big auction which I was looking at yesterday. You've got it on your list.
I think your set belonged to Ian Watson previously then to John Estep.
Peter
Peter: those pictures are my set at an earlier age. The wood of the back
of the main cabinet (not the de novo back insert) was only semi-refinished
and looks similar. There is also visible a defect in the mirror that is an
absolute clincher. The set now is using a new back-surface mirror.
The idea of a metal lever to hold up the lid proved infeasible due to
the size of the little metal tab on the lid.
So I made a large wood block, in a right triangle, originally 6x6x1.5
inches (HxWxThickness) with cut off ends. There is a dug-out spot
that the little metal tab mentioned above sits in, holding the block firmly
in place. This gizmo makes the lid impossible to drop due to bumping.
I'll hijack my own thread ... well, its the same set.
This has been dragged out before on all the other forums, both USA and UK, but not here, so here goes.
The focus evenness is abysmal. The focus control is well in range of getting it right in the center. But, looking down of the tube directly from the front, the upper right is a bit out of focus and the lower left is far out of focus. Turning the focus knob can bring any corner into focus, but the rest of the screen gets worse.
I suspected magnetized works in the set, but a good degaussing had no effect on focus, nor did it move or rotate the image. The yoke is rotated a bit, possibly implying a magnetic effect so I specifically tried various magnets in various places to rotate the image, but that didn't change the focus problem. The scan linearity of this set is completely typical for these things.
Oh yes, this set came with a walk-about coil, which was turned off. Its not needed so I removed it entirely. No change.
There is a bit of wobble in the picture at 10 Hz, caused by stray 60 Hz AC magnetic fields. Its worse on the left side. I can easily buck this out and get a perfectly stable picture, with no change to the focus problem.
Does anybody have another idea to try? This is a problem I've never seen (note: I was never in the TV business, so I've no serious experience in them). I just putz around using the general electronic knowledge I've used extensively in my job as a professor.
I was wondering if the level of EHT would impact on the effects? If it were too low for example.
To understand the black art of electronics is to understand witchcraft. Andrew.
Some first generation EMI TV sets have partial AC coupling between the vision demodulator and the CRT grid or Wehnelt electrode as it was called in those days. For those unfamiliar with these receivers there is no video amplifier following the detector. The CRT has a high slope characteristic allowing only 15volts of positive going video to achieve a fully contrasted picture. Variation in EHT voltage would cause changes to the height and width of the picture. The CRT has a complex hexode gun assembly. Without the circuit diagram to hand I can't remember if the CRT has an accelerator electrode, corresponding to A1 in later tetrode tubes.
Till Eulenspiegel.
All power supply voltages have been checked and there is no significant ripple. This includes the EHT. Note that any effects of power line ripple (EHT is of course off power line) would be at 60 Hz whereas the frame rate is 50 (Or 52.8 for my direct 405 line generator) and so would make obvious fluttery effects. There are none for the focus, and none at all if the AC magnetic field is bucked out.
Could it be possible the sync pulse height is varying? This condition can be caused by excessive RF vision signal level with positive modulation systems. As the RF gain is increased the video part of the modulated waveform hardly increases at all but the height of the sync pulses becomes stretched.
Poor focus can be caused by the resistors in the EHT bleeder chain changing value.
The video demodulator should have 20 volts P - P of positive going video across the diode load resistor.
Till Eulenspiegel.
Do you mean the sync pulse height varying across the picture? The answer is no, I checked using a scope. The problem is the same independant of whether its the Aurora's native color bars (staircase pattern after chopping the color off), program material, or test card C generated natively on my computer. It does not vary with signal level as long as its within sync lock range.
The focus does not vary with "real" time. It does vary as you change the focus control. It does vary as the beam scans, both horizontally and vertically. But the focus voltage at the sync chassis is rock solid. If both line and frame scan voltages got into the focus line somewhere before the tube that would of course explain it. But that could only happen if there is a poor connection between the filter cap on the sync chassis and the electrode inside the tube. There has to be some connection to the 1uF filter cap since the focus control time constant is about a second or a bit more, as expected. The correct average focus voltage is getting to the tube since its in focus somewhere on the screen.
The detector diode output is correct. There's plenty of drive to the tube grid with the contrast set anywhere near the middle, and just a tad too much at the top.
But I will check the connections up to the tube socket, which I am loathe to unplug. What I CAN and will check is the capacitance between the focus electrode and the other electrodes, at its wire connection. It would be low if there is a high resistance between there and the tube innards. That's a good idea! Its the best idea I've found so far.
The picure set you just sent includes a photo of a picture on this very set, long before I got it, but its too small to check focus.
It turns out that I can't do the capacitance test I mentioned because the wires disappear into the unremobavle power chassis.
I did a different test, using a fine crosshatch pattern. Even in the center of the tube there is astigmatism in the beam. The correct focus voltage is different for vertical and horizontal lines. I've never seen this in a magnetic scan tube.
If the astigmatism is fixed and not varying it's safe to say the EHT reservoir capacitor C4 is OK. What's the voltage at the accelerator pin of the CRT? R11 gone high value or OC. C6 leaky or SC?
Till Eulenspiegel.
The problem I have with my Marconi 702 is poor contrast. The output from the vision demodulator is a best only 15volts including syncs. Turn up the sensitivity control and the effect mentioned in a previous post is sync pulse stretching and little change in the amplitude of the video part of the waveform.
Till Eulenspiegel.
Till: (I keep hearing your theme song, especially that last low note)
My set has all the voltages quite OK.
As to low video, I presume you've tried everything and the voltages are OK.
Can you turn up the RF at its source? I'm using a rather high attenuation attenuator. One can get more output with bigger signal but it indeed starts compressing and loses sync. I assume you've tried different tubes, especially the last amplifier tube, the MSP41.
The only other thing I can think of is bad alignment. The RF gain is of course
done using the very unusual way of changing screen voltages. I found that there was one alignment situation where the gain was too high early and it clipped in one of the intermediate stages. This was done with a 100 MHz scope
not using a probe but rather a plain piece of RG58 to the scope and a mini alligator clip clipped to the insulation of the RF-bearing wires, making a capacitative voltage divider with about 0.5pF capitance (into one megohm).
I find it very very hard to believe that with good tubes and correct voltages
at modest to high gain you can't find an alignment that gives good voltage.
As I've said before, I found alignment a true nightmare, and the instructions
failed abysmally. I eventually settled on using only the lower sideband
as VSB and a fairly wide video bandwidth.
Given the astigmatism, I lead to think that maybe an electrode is askew in the tube, given the otherwise normal behavior.
